Countering cyberbullying in social networks

The study is devoted to the development of a program for determining the text tonality. The paper substantiates the relevance of protecting society from cyberbullying. The methods of cyberbullying countering are analyzed. An indicator of the negativity of the site's information was introduced. The work of the site blocker is considered in detail. The use of sentiment analysis, which is based on the use of neural networks, is justified. For the sentiment analysis of information flows, a program has been developed in the high-level programming language Python based of ready-made trained neural networks. The stem dictionary is used. Information flows are divided into tokens represented as vectors. In detail, the examples show the use of various neural networks to determine the tonality of the text. The results of the two text analysis codes are compared using the probability of obtaining the correct level of negativity of the text. The expediency of using site blockers as methods to protect against cyberbullying as well as the use of datasets for training neural networks are justified.